[…]DUI BLOG: High Breathalyzer Readings….from Dieting[…]…

Essentially, different functional groups and bonds of a molecule will absorb different frequencies of IR.

Any given molecule can have any number of functional groups and bonds. Ethanol, for example, has a C-C bond, a C-O bond, C-H bonds and an O-H bond.

The ranges of frequencies that get absorbed by these bonds do overlap. In fact, the range for O-H is pretty big and overlaps with a number of substances.

So, we have two problems if these devices look for O-H bonds:
1. Substances other than alcohol can have the same O-H bonds, such as the sugar alcohols in artificial sweeteners
2. Other bonds/groups can and do overlap with the frequencies absorbing O-H

Regarding point #2, different groups may have more or less of an impact as well, so small amounts of an “interferent” could have large effects.

Overall, IR Spectroscopy is fast and cheap, but not too reliable, _especially_ when there are multiple substances at hand.

Since there’s a variety of ways an IR spectroscopy algorithm could look to see if alcohol is present or not and in what quantity, it’s entirely possible that different detectors will be fooled very differently by these “interferents”.

Any organic chemist would agree that IR Spectroscopy of a mixture of multiple substances can be quite useless when you want to determine how much of a substance is present. It’s not too bad for identifying what is present, but when we’re dealing with ethanol, we need to know how much is present too.

And like you said, many factors can influence how much ethanol comes out of the lungs. If one has a fever, that can increase the amount of alcohol that gets vaporized.